Permanent Multi-Well DAS Monitoring Explains The Seismicity Triggered by an Interaction Between Two CO2 Plumes

Induced seismicity is often perceived as one of the main risks for geological storage of carbon dioxide. Therefore, microseismic monitoring is a necessary component for the storage projects. However, only three dedicated CO2 injections into saline aquifers have led to detectable microseismic events, with the largest documented event having a moment magnitude of 1.6. This study presents another field experiment, CO2CRC Otway Project Stage 3, where two small injections of supercritical CO2 (15,000 tonnes and 0.2 MPa pressure) into a good sandstone aquifer induced detectable seismicity.

In 610 days passive seismic monitoring of the Stage 3 injection we detected 17 microseismic events (maximum moment magnitude MW0.1) using 5 deep boreholes with engineered optical fiber. Overall, we found that the DAS array provides sufficient sensitivity and signal-to-noise ratio to warrant detection of induced events with ~MW-2 that occurred up to a kilometre away from the nearest borehole. The main cluster of the events has the same location and source mechanism as the one triggered by a previous twin injection at the Otway Project site, Stage 2C. Also, timing for the Stage 2C and Stage 3 events is closely correlated with the actual movement of the CO2 plume front rather than the typically observed situation, where events are triggered by the pore pressure plume. The nature of the plume-fault interaction remains unclear yet, but some mechanisms such as chemical alteration of the fault gouge may enhance the susceptibility of the fault to pressure perturbation for a very small injection or a leakage. Moreover, the Stage 3 monitoring program included an extensive pressure tomography. The associated high-pressure but short brine injections led to some detectable seismicity outside of the CO2 plume, with no obvious clustering of the hypocenters along some subseismic faults.

It is important to note that the seismogenic fault could not be identified in the seismic images prior to the injection. Only through the high-precision tracking of the two injections using the DAS array we could detect and characterize the reactivated fault.